OECD/OCDE TG 431 Adopted:

26 September 2014

1

© OECD, (2014)

You are free to use this material for personal, non-commercial purposes without seeking prior consent

from the OECD, provided the source is duly mentioned. Any commercial use of this material is subject to

written permission from the OECD.

OECD GUIDELINES FOR THE TESTING OF CHEMICALS

IN VITRO SKIN CORROSION: RECONSTRUCTED HUMAN EPIDERMIS (RHE) TEST

METHOD

INTRODUCTION

1. Skin corrosion refers to the production of irreversible damage to the skin manifested as visible

necrosis through the epidermis and into the dermis, following the application of a test chemical [as defined

by the United Nations (UN) Globally Harmonized System of Classification and Labelling of Chemicals

(GHS)] (1). This updated Test Guideline 431 provides an in vitro procedure allowing the identification of

non-corrosive and corrosive substances and mixtures in accordance with UN GHS (1). It also allows a

partial sub-categorization of corrosives.

2. The assessment of skin corrosivity has typically involved the use of laboratory animals (OECD

Test Guideline 404 (TG 404); adopted in 1981 and revised in 1992 and 2002) (2). In relation to animal

welfare concerns, TG 404 recommends the use of a tiered testing strategy for the determination of skin

corrosion and irritation which includes the use of validated in vitro or ex vivo test methods avoiding pain

and suffering of animals. In addition to TG 431 (originally adopted in 2004) (3), two other in vitro test

methods for testing of corrosivity have been validated and adopted as OECD Test Guidelines 430 (4) and

435 (5). Three validated in vitro test methods have been adopted as OECD TG 439 (6), to be used for the

skin irritation part of the tiered testing strategy of TG 404 (2).

3. This Test Guideline addresses the human health endpoint skin corrosion. It makes use of

reconstructed human epidermis (RhE) (obtained from human derived non-transformed epidermal

keratinocytes) which closely mimics the histological, morphological, biochemical and physiological

properties of the upper parts of the human skin, i.e. the epidermis. This Test Guideline was originally

adopted in 2004 and updated in 2013 and 2014 to include a set of Performance Standards (PS) (Annex 1)

for the assessment of similar and modified RhE-based test methods (7), in accordance with the principles

of Guidance Document No. 34 (8). Other updates comprise the addition of two test methods using the RhE

TG 431 OECD/OCDE

2

© OECD, (2014)

models SkinEthic™ RHE1 and epiCS

® (previously named EST-1000), and the possibility to use the

methods to support the sub-categorisation of corrosive chemicals.

4. Four validated test methods using commercially available RhE models are included in this Test

Guideline. Prevalidation studies (9), followed by a formal validation study for assessing skin corrosion

(10)(11) (12) have been conducted (13) (14) for two of these commercially available test methods,

EpiSkin™ Standard Model (SM) and EpiDerm™ Skin Corrosivity Test (SCT) (EPI-200) (referred to in the

following text as the Validated Reference Methods – VRMs). The outcome of these studies led to the

recommendation that the two VRMs mentioned above could be used for regulatory purposes for

distinguishing corrosive (C) from non-corrosive (NC) substances, and that the EpiSkin™ could moreover

be used to support sub-categorization of corrosive substances (15) (16) (17). Two other commercially

available in vitro skin corrosion RhE test methods have shown similar results to the EpiDerm™ VRM

according to PS-based validation (18) (19) (20). These are the SkinEthic™ RHE and epiCS® (previously

named EST-1000) that can also be used for regulatory purposes for distinguishing corrosive from non-

corrosive substances (21) (22). Post validation studies performed by the RhE model producers in the years

2012 to 2014 with a refined protocol correcting interferences of unspecific MTT reduction by the test

chemicals improved the performance of both discrimination of C/NC as well as supporting sub-

categorisation of corrosives (23) (24).

5. Before a proposed similar or modified in vitro RhE test method for skin corrosion other than the

VRMs can be used for regulatory purposes, its reliability, relevance (accuracy), and limitations for its

proposed use should be determined to ensure its similarity to the VRMs, in accordance with the

requirements of the PS set out in this Test Guideline (Annex 1). The Mutual Acceptance of Data will only

be guaranteed after any proposed new or updated test method following the PS of this Test Guideline have

been reviewed and included in this Test Guideline. The test methods included in this Test Guideline can be

used to address countries’ requirements for test results on in vitro test method for skin corrosion,while

benefiting from the Mutual Acceptance of Data.

DEFINITIONS

6. Definitions used are provided in Annex 2.

INITIAL CONSIDERATIONS

7. This Test Guideline addresses the in vitro skin corrosion component of the tiered testing strategy

recommended within TG 404 for dermal corrosion/irritation assessment (2) (25). It allows the

identification of non-corrosive and corrosive substances and mixtures in accordance with the UN GHS (1).

This Test Guideline further supports the sub-categorization of corrosive substances and mixtures into

optional Category 1A, in accordance with the UN GHS (1), as well as a combination of Categories 1B and

1C (23) (24). A limitation of this Test Guideline is that it does not allow discriminating between skin

corrosive sub-categories 1B and 1C in accordance with the UN GHS (1) due to the limited set of well-

1 Please note that the abbreviation RhE (=Reconstructed human Epidermis) is used for all models based on RhE

technology. The abbreviation RHE as used in conjunction with the SkinEthicTM

model means the same, but, as

part of the name of this specific test method as marketed, is spelled all in capitals.

OECD/OCDE TG 431

3

© OECD, (2014)

known in vivo corrosive Category 1C chemicals. EpiSkinTM

, EpiDermTM

, SkinEthicTM

and epiCS®test

methods are able to sub-categorize (i.e. 1A versus 1B-and-1C versus NC) but differences are observed

between EpiSkinTM

and the three other test methods, EpiDermTM

, SkinEthicTM

and epiCS® in view of their

capacity to provide information on sub-categorisation. Results from EpiSkinTM

can be used as such;

whereas results from EpiDermTM

, SkinEthicTM

and epiCS® generate high over-classification rates for a

combination of categories 1B and 1C (see Annex 4). Therefore, for EpiDermTM

, SkinEthicTM

and epiCS®,

chemicals that are classified as 1B-and-1C can be considered as 1B-and-1C, while chemicals for which cell

viability at 3 minutes is below 50% should just be considered as Category 1, since the Category 1A

predictions of these three test methods contain a high rate of over-predictions of chemicals of Categories

1B and 1C. The regulatory framework in member countries will decide how this Test Guideline will be

used, e.g. acknowledging the significant probability of overclassification, a Category 1A classification may

still be accepted or further testing may be conducted to confirm the result.

8. A wide range of chemicals representing mainly individual substances has been tested in the

validation supporting the test methods included in this Test Guideline when they are used for identification

of non-corrosives and corrosives; the empirical database of the validation study amounted to 60 chemicals

covering a wide range of chemical classes (10) (11) (12). Testing to demonstrate sensitivity, specificity,

accuracy and within-laboratory-reproducibility of the assay for sub-categorization was performed by the

test method developers and results were reviewed by the OECD (23) (24). On the basis of the overall data

available, the Test Guideline is applicable to a wide range of chemical classes and physical states including

liquids, semi-solids, solids and waxes. The liquids may be aqueous or non-aqueous; solids may be soluble

or insoluble in water. Whenever possible, solids should be ground to a fine powder before application; no

other prior treatment of the sample is required. In cases where evidence can be demonstrated on the non-

applicability of test methods included in the Test Guideline to a specific category of test chemicals, these

test methods should not be used for that specific category of test chemicals. In addition, this Test Guideline

is assumed to be applicable to mixtures as an extension of its applicability to substances. However, due to

the fact that mixtures cover a wide spectrum of categories and composition, and that only limited

information is currently available in the public domain on the testing of mixtures, in cases where evidence

can be demonstrated on the non-applicability of the Test Guideline to a specific category of mixtures

(e.g. following a strategy as proposed in (26)), the Test Guideline should not be used for that specific

category of mixtures. Gases and aerosols have not been assessed yet in validation studies (10) (11) (12).

While it is conceivable that these can be tested using RhE technology, the current Test Guideline does not

allow testing of gases and aerosols. It should also be noted that some chemicals may interfere with the cell

viability measurements and need the use of adapted controls for corrections (see paragraphs 25 to 27).

9. While this Test Guideline does not provide adequate information on skin irritation, it should be

noted that OECD TG 439 specifically addresses the health effect skin irritation in vitro and is based on the

same RhE test system, though using another protocol (6). For a full evaluation of local skin effects after a

single dermal exposure, it is recommended to follow the sequential testing strategy as appended to TG 404

(2) (25). This testing strategy includes the conduct of in vitro tests for skin corrosion (such as described in

this Test Guideline) and skin irritation before considering testing in living animals. It is recognized that the

use of human skin is subject to national and international ethical considerations and conditions.

TG 431 OECD/OCDE

4

© OECD, (2014)

10. This Test Guideline also includes a set of Performance Standards (PS) (Annex 1) for determining

the validation status (reliability and relevance) of similar and modified skin corrosion test methods that are

structurally and mechanistically similar to the VRMs (7), in accordance with the principles of Guidance

Document No. 34 (8). These PS include a list of Reference Substances by which to evaluate assay

performance, the essential test method components by which to evaluate the structural, mechanistic and

procedural similarity of a new proposed test method, and the minimum reliability and accuracy values

necessary for the test method to be considered comparable to the VRMs. Within the Reference Chemical

list, a subset of 13 Proficiency Substances (Table 1) is provided to be used by laboratories to demonstrate

proficiency in using in vitro human skin models (see paragraphs 13 and 14).

PRINCIPLE OF THE TEST

11. The test chemical is applied topically to a three-dimensional RhE model, comprised of non-

transformed, human-derived epidermal keratinocytes, which have been cultured to form a multi-layered,

highly differentiated model of the human epidermis. It consists of organized basal, spinous and granular

layers, and a multi-layered stratum corneum containing intercellular lamellar lipid layers representing main

lipid classes analogous to those found in vivo.

12. The RhE test method is based on the premise that corrosive chemicals are able to penetrate the

stratum corneum by diffusion or erosion, and are cytotoxic to the cells in the underlying layers. Cell

viability is measured by enzymatic conversion of the vital dye MTT [3-(4,5-Dimethylthiazol-2-yl)-2,5-

diphenyltetrazolium bromide, Thiazolyl blue tetrazolium bromide; CAS number 298-93-1], into a blue

formazan salt that is quantitatively measured after extraction from tissues (27). Corrosive chemicals are

identified by their ability to decrease cell viability below defined threshold levels (see paragraphs 31 and

32). The RhE-based skin corrosion test methods have shown to be predictive of in vivo skin corrosion

effects assessed in rabbits according to the OECD guideline 404 (2).

DEMONSTRATION OF PROFICIENCY

13. Prior to routine use of any of the four validated RhE test methods that adhere to this Test

Guideline, laboratories should demonstrate technical proficiency by correctly classifying the twelve

Proficiency Substances listed in Table 1. In case of the use of a method for sub-classification, also the

correct sub- categorization should be demonstrated.

OECD/OCDE TG 431

5

© OECD, (2014)

Table 1: List of Proficiency Substances

Chemical1 CASRN Chemical Class

2

UN GHS

Cat. Based

on In Vivo

results 3

VRM

Cat. Based

on In Vitro

results4

MTT

Reducer5

Physical

State

Category 1A In Vivo Corrosives

Bromoacetic acid 79-08-3 Organic acid 1A (3) 1A -- S

Boron trifluoride

dihydrate 13319-75-0 Inorganic acid 1A (3) 1A -- L

Phenol 108-95-2 Phenol 1A (3) 1A -- S

Dichloroacetyl

chloride 79-36-7 Electrophile 1A (3) 1A -- L

Category 1B/1C In Vivo Corrosives

Glyoxylic acid

monohydrate 563-96-2 Organic acid 1B/1C (3) 1B/1C -- S

Lactic acid 598-82-3 Organic acid 1B/1C (3) 1B/1C -- L

Ethanolamine 141-43-5 Organic base 1B (3) 1B/1C Y Viscous

Hydrochloric acid

(14.4%) 7647-01-0 Inorganic acid 1B/1C (3) 1B/1C -- L

In Vivo Non Corrosives

Phenethyl bromide 103-63-9 Electrophile NC (3) NC Y L

4-Amino-1,2,4-

triazole 584-13-4 Organic base NC (3) NC -- S

4-(methylthio)-

benzaldehyde 3446-89-7 Electrophile NC (3) NC Y L

Lauric acid 143-07-7 Organic acid NC (3) NC -- S

Abbreviations: CASRN = Chemical Abstracts Service Registry Number; UN GHS = United Nations Globally Harmonized System

(1); VRM = Validated Reference Method; NC = Not Corrosive 1These substances, sorted first by corrosives versus non-corrosives, then by corrosive sub-category and then by chemical class,

were selected from the substances used in the ECVAM validation studies of EpiSkin™ and EpiDermTM (10) (11) (12) and from

post-validation studies based on data provided by EpiSkinTM (24), EpiDermTM, SkinEthicTM and epiCS® developers. Unless

otherwise indicated, the substances were tested at the purity level obtained when purchased from a commercial source (10) (12).

The selection includes, to the extent possible, substances that: (i) are representative of the range of corrosivity responses (e.g. non-

corrosives; weak to strong corrosives) that the VRMs are capable of measuring or predicting; (ii) are representative of the chemical

classes used in the validation studies; (iii) have chemical structures that are well-defined; (iv) induce reproducible results in the

VRM; (v) induce definitive results in the in vivo reference test method; (vi) are commercially available; and (vii) are not associated

with prohibitive disposal costs. 2Chemical class assigned by Barratt et al. (1998) (10). 3The corresponding UN Packing groups are I, II and III, respectively, for the UN GHS 1A, 1B and 1C. 4The VRM in vitro predictions reported in this table were obtained with the EpiSkinTM and the EpiDermTM test methods (VRMs)

during post-validation testing performed by the test method developers. 5The viability values obtained in the ECVAM Skin Corrosion Validation Studies were not corrected for direct MTT reduction

(killed controls were not performed in the validation studies). However, the post-validation data generated by the test method

developers that are presented in this table were acquired with adapted controls.

TG 431 OECD/OCDE

6

© OECD, (2014)

14. As part of the proficiency exercise, it is recommended that the user verifies the barrier properties

of the tissues after receipt as specified by the RhE model manufacturer. This is particularly important if

tissues are shipped over long distance/time periods. Once a test method has been successfully established

and proficiency in its use has been demonstrated, such verification will not be necessary on a routine basis.

However, when using a test method routinely, it is recommended to continue to assess the barrier

properties in regular intervals.

PROCEDURE

15. The following is a generic description of the components and procedures of the RhE test methods

for skin corrosion assessment covered by this Test Guideline. The RhE models endorsed as scientifically

valid for use within this Test Guideline, i.e. the EpiSkinTM

(SM), EpiDerm™ (EPI-200), SkinEthicTM

RHE

and epiCS®

models (18) (19) (20) (28) (29) (30) (31) (32) (33), can be obtained from commercial sources.

Standard Operating Procedures (SOPs) for these four RhE models are available (34) (35) (36) (37), and

their main test method components are summarized in Annex 3. It is recommended that the relevant SOP

be consulted when implementing and using one of these methods in the laboratory. Testing with the four

RhE test methods covered by this Test Guideline should comply with the following:

RHE TEST METHOD COMPONENTS

General Conditions

16. Non-transformed human keratinocytes should be used to reconstruct the epithelium. Multiple

layers of viable epithelial cells (basal layer, stratum spinosum, stratum granulosum) should be present

under a functional stratum corneum. The stratum corneum should be multi-layered containing the essential

lipid profile to produce a functional barrier with robustness to resist rapid penetration of cytotoxic marker

chemicals, e.g. sodium dodecyl sulfate (SDS) or Triton X-100. The barrier function should be

demonstrated and may be assessed either by determination of the concentration at which a marker chemical

reduces the viability of the tissues by 50% (IC50) after a fixed exposure time, or by determination of the

exposure time required to reduce cell viability by 50% (ET50) upon application of the marker chemical at a

specified, fixed concentration (see paragraph 18). The containment properties of the RhE model should

prevent the passage of material around the stratum corneum to the viable tissue, which would lead to poor

modelling of skin exposure. The RhE model should be free of contamination by bacteria, viruses,

mycoplasma, or fungi.

Functional Conditions

Viability

17. The assay used for determining the magnitude of viability is the MTT-assay (27). The RhE model

users should ensure that each batch of the RhE model used meets defined criteria for the negative control.

The optical density (OD) of the extraction solvent alone should be sufficiently small, i.e. OD < 0.1. An

acceptability range (upper and lower limit) for the negative control OD values is established by the RhE

model developer/supplier, and the acceptability ranges for the four validated RhE test methods included in

this Test Guideline are given in Table 2. It should be documented that the tissues treated with negative

control are stable in culture (provide similar OD measurements) for the duration of the exposure period.

OECD/OCDE TG 431

7

© OECD, (2014)

Table 2: Acceptability ranges for negative control OD values to control batch quality

Lower acceptance limit Upper acceptance limit

EpiSkin™ (SM) ≥ 0.6 ≤ 1.5

EpiDerm™ SCT (EPI-200) ≥ 0.8 ≤ 2.8

SkinEthicTM

RHE ≥ 0.8 ≤ 3.0

epiCS®

≥ 0.8 ≤ 2.8

Barrier function

18. The stratum corneum and its lipid composition should be sufficient to resist the rapid penetration

of certain cytotoxic marker chemicals (e.g. SDS or Triton X-100), as estimated by IC50 or ET50

(see paragraph 21).

Morphology

19. Histological examination of the RhE model should be performed demonstrating multi-layered

human epidermis-like structure (containing stratum basale, stratum spinosum, stratum granulosum and

stratum corneum and exhibits lipid profile similar to lipid profile of human epidermis.

Reproducibility

20. Test method users should demonstrate reproducibility of the test methods over time with the

positive and negative controls. Furthermore, the test method should only be used if the RhE model

developer/supplier provides data demonstrating reproducibility over time with corrosive and non-corrosive

chemicals from e.g. the list of Proficiency Substances (Table 1). In case of the use of a method for sub-

categorization, also the reproducibility of sub-categorization should be demonstrated.

Quality control (QC)

21. The RhE model should only be used if the developer/supplier demonstrates that each batch of the

RhE model used meets defined production release criteria, among which those for viability (paragraph 17),

barrier function (paragraph 18) and morphology (paragraph 19) are the most relevant. These data are

provided to the test method users, so that they are able to include this information in the test report. Only

results produced with QC accepted tissue batches can be accepted for reliable prediction of corrosive

classification. An acceptability range (upper and lower limit) for the IC50 or the ET50 is established by the

RhE model developer/supplier. The acceptability ranges for the four validated test methods are given in

Table 3.

TG 431 OECD/OCDE

8

© OECD, (2014)

Table 3:QC batch release criteria

Lower acceptance limit Upper acceptance limit

EpiSkinTM

(SM)

(18 hours treatment with

SDS)(35)

IC50 = 1.0 mg/mL IC50 = 3.0 mg/mL

EpiDerm™ SCT (EPI-200)

(1% Triton X-100)(36)

ET50 = 4.0 hours ET50 = 8.7 hours

SkinEthicTM

RHE

(1% Triton X-100)(37)

ET50 = 4.0 hours ET50 = 10.0 hours

epiCS®(1% Triton X-100)(38) ET50 = 2.0 hours ET50 = 7.0 hours

Application of the Test Chemical and Control Substance

22. At least two tissue replicates should be used for each test chemical and controls for each exposure

time. For liquid as well as solid chemicals, sufficient amount of test chemical should be applied to

uniformly cover the epidermis surface while avoiding an infinite dose, i.e. a minimum of 70 μL/cm2 or 30

mg/cm2 should be used. Depending on the methods, the epidermis surface should be moistened with

deionized or distilled water before application of solid chemicals, to improve contact between the test

chemical and the epidermis surface (34) (35) (36) (37). Whenever possible, solids should be tested as a fine

powder. The application method should be appropriate for the test chemical (see e.g. references 12, 35-

38). At the end of the exposure period, the test chemical should be carefully washed from the epidermis

with an aqueous buffer, or 0.9% NaCl. Depending on which of the four validated RhE test methods is used,

two or three exposure periods are used per test chemical (for all four valid RhE models: 3 min and 1 hour;

for EpiSkinTM

an additional exposure time of 4 hours). Depending on the RhE test method used and the

exposure period assessed, the incubation temperature during exposure may vary between room temperature

and 37ºC.

23. Concurrent negative and positive controls (PC) should be used in each run to demonstrate that

viability (with negative controls), barrier function and resulting tissue sensitivity (with the PC) of the

tissues are within a defined historical acceptance range. The suggested PC chemicals are glacial acetic acid

or 8N KOH depending upon the RhE model used. It should be noted that 8N KOH is a direct MTT reducer

that might require adapted controls as described in paragraphs 25 and 26. The suggested negative controls

are 0.9% (w/v) NaCl or water.

Cell Viability Measurements

24. The MTT assay, which is a quantitative assay, should be used to measure cell viability under this

Test Guideline (28). The tissue sample is placed in MTT solution of appropriate concentration (0.3 or

1 mg/mL) for 3 hours. The precipitated blue formazan product is then extracted from the tissue using a

solvent (e.g. isopropanol, acidic isopropanol), and the concentration of formazan is measured by

determining the OD at 570 nm using a filter band pass of maximum ± 30 nm.

OECD/OCDE TG 431

9

© OECD, (2014)

25. Test chemicals may interfere with the MTT assay, either by direct reduction of the MTT into blue

formazan, and/or by colour interference if the test absorbs, naturally or due to treatment procedures, in the

same OD range of formazan (570 ± 30 nm, mainly blue and purple chemicals). Additional controls should

be used to detect and correct for a potential interference from these test chemicals (see paragraphs 26 and

27). This is especially important when a specific test chemical is not completely removed from the tissue

by rinsing or when it penetrates the epidermis, and is therefore present in the tissues when the MTT

viability test is performed. Detailed description of how to correct direct MTT reduction and interferences

by colouring agents is available in the SOPs for the test methods (34) (35) (36) (37). Non-specific MTT

reduction (NSMTT) and non-specific colour (NSC) due to these interferences may increase the OD of the

tissue extract above the linearity range of the spectrophotometer. It is therefore important for each

laboratory to determine the OD linearity range of their spectrophotometer with e.g. MTT formazan

(CAS # 57360-69-7) commercially available from Sigma (Ref: M2003) before initiating the testing of test

chemicals for regulatory purposes. When the OD of the tissue extract falls above the linearity range of the

spectrophotometer, it should be diluted in acidified isopropanol or acidic isopropanol and the dilution

factor should be taken into account when determining % NSMTT and/or % NSC relative to the negative

controls ran concurrently to the test being corrected. Test results for materials inducing %NSMTT and/or

%NSC 50% of negative control should be taken with caution.

26. To identify direct MTT reducers, each test chemical should be added to freshly prepared MTT

medium. The mixture is incubated in the dark at 37°C, 5% CO2 for a minimum of 60 min and a maximum

of 180 min depending upon the RhE model used (34) (35) (36) (37). MTT medium is used as control. If the

MTT mixture containing the test chemical (or suspension for insoluble compounds) turns blue/purple, the

test chemical is presumed to directly reduce the MTT, and further functional check on non-viable

epidermis should be performed. This additional functional check employs killed tissues that possess no

metabolic activity but absorb and bind the test chemical in similar amount as viable tissues. Each MTT

reducing chemical is applied on at least two killed tissue replicates per exposure time, which undergo the

whole skin corrosion test. True tissue viability is calculated as the difference between the OD obtained

with living tissue treated by MTT reducer and the OD obtained with frozen tissue treated by MTT reducer,

and subsequently divided by the OD of the Negative Control concurrently tested.

27. To identify colour interference, spectral analysis of a coloured chemical in water (environment

during exposure) and/or isopropanol (extracting solution) should be performed to evaluate if the chemical

requires additional controls. If the chemical in water and/or isopropanol absorbs light in the range of

570 ± 30 nm, colorant controls should be performed. Each coloured chemical is applied on at least two

viable tissue replicates per exposure time, which undergo the entire skin corrosion test but are incubated

with medium instead of MTT solution during the MTT incubation step. An independent NSC control needs

to be performed concurrently per exposure time per coloured test chemical (in each run) due to the inherent

biological variability of living tissues. True tissue viability is calculated as the difference between the OD

obtained with living tissues incubated with MTT solution and the OD obtained with living tissues

incubated with medium without MTT, and subsequently divided by the OD of the Negative Control

performed in the same run.

TG 431 OECD/OCDE

10

© OECD, (2014)

Acceptability Criteria

28. For each test method using valid RhE models, tissues treated with the negative control should

exhibit OD reflecting the quality of the tissues as described in table 2 and should not be below historically

established boundaries. Tissues treated with the PC, i.e. glacial acetic acid or 8N KOH, should reflect the

ability of the tissues to respond to a corrosive chemical under the conditions of the test method (see Annex

3). The variability between tissue replicates of test and/or control chemicals should fall within the accepted

limits for each valid RhE model requirements (see Annex 3) (e.g. the difference of viability between the

two tissue replicates should not exceed 30%). If either the negative control or PC included in a run fall out

of the accepted ranges, the run is considered as not qualified and should be repeated. If the variability of

test chemicals falls outside of the defined range, its testing should be repeated.

Interpretation of Results and Prediction Model

29. The OD values obtained for each test chemical should be used to calculate percentage of viability

relative to the negative control, which is set at 100%. The cut-off percentage cell viability values

distinguishing corrosive from non-corrosive test chemical (or discriminating between different corrosive

sub-categories) are defined below in paragraphs 31 and 32 for each of the test methods covered by this

Test Guideline and should be used for interpreting the results.

30. A single testing run composed of at least two tissue replicates should be sufficient for a test

chemical when the resulting classification is unequivocal. However, in cases of borderline results, such as

non-concordant replicate measurements, a second run may be considered, as well as a third one in case of

discordant results between the first two runs.

31. The prediction model for the EpiSkin skin corrosion test method (11) (34) (24), associated with

the UN GHS (1) classification system, is shown in Table 4:

Table 4:EpiSkinTM

prediction model

Viability measured after exposure time

points (t=3, 60 and 240 minutes)

Prediction

to be considered

< 35% after 3 min exposure Corrosive:

Optional Sub-category 1A *

≥ 35% after 3 min exposure AND

< 35% after 60 min exposure

OR

≥ 35% after 60 min exposure AND

< 35% after 240 min exposure

Corrosive:

A combination of optional

Sub-categories 1B and 1C

≥ 35% after 240 min exposure Non-corrosive

*) According to the data generated in view of assessing the usefulness of the RhE test methods for supporting sub-

categorisation, it was shown that around 22% of the Cat1A results of the EpiSkinTM

test method may actually

constitute Category 1B or 1C substances/mixtures (i.e. over classifications) (see Table 4.0 in Annex 4).

OECD/OCDE TG 431

11

© OECD, (2014)

32. The prediction models for the EpiDerm SCT (13) (36), the SkinEthicTM

RHE (19) (20) (36),

and the epiCS® (18) (37) skin corrosion test methods, associated with the UN GHS (1) classification

system, are shown in Table 5:

Table 5:EpiDermTM

SCT, SkinEthicTM

RHE and epiCS®

Viability measured after exposure time

points (t=3 and 60 minutes)

Prediction

to be considered

< 50% after 3 min exposure Corrosive:

Optional Sub-category 1A*

≥ 50% after 3 min exposure AND

< 15% after 60 min exposure

Corrosive:

A combination of optional

Sub-categories 1B and 1C

≥ 50% after 3 min exposure AND

≥ 15% after 60 min exposure Non-corrosive

*) According to the data generated in view of assessing the usefulness of the RhE test methods for supporting sub-

categorisation, it was shown that around 42% of the Cat1A results of the EpiDermTM

test method, and around 46% of

the Cat 1A results of the SkinEthicTM

and the epiCS®

test method may actually constitute Category 1B or 1C

substances/mixtures (i.e. over-classifications) (see Table 4.0 in Annex 4).

DATA AND REPORTING

Data

33. For each test, data from individual tissue replicates (e.g. OD values and calculated percentage

cell viability for each test chemical, including classification) should be reported in tabular form, including

data from repeat experiments as appropriate. In addition, means and ranges of viability and CVs between

tissue replicates for each test should be reported. Observed interactions with MTT reagent by direct MTT

reducers or coloured test chemicals should be reported for each tested chemical.

Test report

34. The test report should include the following information:

Test Chemical and Control Substance:

– Chemical name(s) such as IUPAC or CAS name and number, if known;

– Purity and composition of the substance or mixture (in percentage(s) by weight);

– Physical-chemical properties relevant to the conduct of the study (e.g. physical state, stability,

volatility, pH, water solubility, if known);

– Treatment of the test chemical /control substance prior to testing, if applicable (e.g. warming,

grinding);

– Storage conditions;

TG 431 OECD/OCDE

12

© OECD, (2014)

RhE model and protocol used and rationale for it (if applicable)

Test Conditions:

– RhE model used (including batch number);

– Calibration information for measuring device (e.g. spectrophotometer), band pass used for

measuring cell viability, and OD linearity range of measuring device;

– Complete supporting information for the specific RhE model used including its performance.

This should include, but is not limited to:

i) Viability

ii) Barrier function

iii) Morphology

iv) Reproducibility and predictive capacity

v) Quality controls (QC) of the model

– Details of the test procedure used. This should include, but is not limited to;

i) Washing procedures after exposure period

ii) Wavelength and band pass (filter) used to measure OD (cell viability)

– Test doses used, duration of exposure period(s) and temperature(s) of exposure;

– Number of tissue replicates used per test chemical and controls (PC, negative control, and

NSMTT and NSC, if applicable), per exposure time;

– Indication of controls used for direct MTT-reducers and/or colouring test chemicals;

– Description of any modifications of the test procedure (including washing procedures);

– Reference to historical data of the model. This should include, but is not limited to;

i) Acceptability of the QC data with reference to historical data

ii) Acceptability of the positive and negative control values with reference to positive

and negative control means and ranges

iii) Acceptability of the test results with reference to historical variability between

tissue replicates

– Description of decision criteria/prediction model applied based on the RhE model used;

Results:

– Tabulation of data for individual test chemicals and controls, for each exposure period, each

run and each replicate measurement, means, ranges and CVs, and the derived

classification;

– Results of controls used for direct MTT-reducers and/or colouring test chemicals;

– Description of other effects observed;

– The derived classification with reference to the prediction model/decision criteria used;

Discussion of the results

Conclusions

OECD/OCDE TG 431

13

© OECD, (2014)

LITERATURE

(1) UN (2013), United Nations Globally Harmonized System of Classification and Labelling of

Chemicals (GHS), Fifth revised edition, UN New York and Geneva.

Available at: http://www.unece.org/trans/danger/publi/ghs/ghs_rev05/05files_e.html.

(2) OECD (2002), Test No. 404: Acute Dermal Irritation/Corrosion, OECD Guideline for the Testing of

Chemicals, Section 4, OECD Publishing, Paris. http://dx.doi.org/10.1787/9789264070622-en.

(3) OECD (2004), Test No. 431: In vitro skin model, OECD Guideline for the Testing of Chemicals,

Section 4, OECD Publishing, Paris. http://dx.doi.org/10.1787/9789264071148-en.

(4) OECD (2004), Test No. 430: In vitro skin corrosion: Transcutaneous Electrical Resistance (TER),

OECD Guideline for the Testing of Chemicals, Section 4, OECD Publishing, Paris.

http://dx.doi.org/10.1787/9789264071124-en.

(5) OECD (2006), Test No. 435: In vitro membrane barrier test method, OECD Guideline for the

Testing of Chemicals, Section 4, OECD Publishing, Paris.

http://dx.doi.org/10.1787/9789264067318-en.

(6) OECD (2010), Test No. 439: In vitro skin irritation: reconstructed human epidermis test method,

OECD Guideline for the Testing of Chemicals, Section 4, OECD Publishing, Paris.

http://dx.doi.org/10.1787/9789264090958-en.

(7) ICCVAM (2004), Recommended Performance Standards for In Vitro Test Methods for Skin

Corrosion, NIH Publication Number 04-4510, Research Triangle Park, NC: National Institute of

Environmental Health Sciences, http://iccvam.niehs.nih.gov/docs/dermal_docs/ps/ps044510.pdf .

(8) OECD (2005), “Guidance Document on the Validation and International Acceptance of New or

Updated Test Methods for Hazard Assessment”, OECD Environment, Health and Safety

Publications (EHS), Series on Testing and Assessment, No. 34, OECD Publishing, Paris.

(9) Botham, P.A. et al. (1995), A prevalidation study on in vitro skin corrosivity testing: The report and

recommendations of ECVAM Workshop 6, ATLA, Vol. 23, pp. 219-255.

(10) Barratt, M.D. et al. (1998), The ECVAM international validation study on in vitro tests for skin

corrosivity. 1. Selection and distribution of the test chemicals, Toxicology in Vitro, Vol. 12/4, pp.

471-482.

(11) Fentem, J.H. et al. (1998), The ECVAM international validation study on in vitro tests for skin

corrosivity. 2. Results and evaluation by the Management Team, Toxicology in Vitro, 12/4, pp. 483-

524.

(12) Liebsch, M. et al. (2000), The ECVAM prevalidation study on the use of EpiDerm for skin

corrosivity testing, ATLA, Vol. 28, pp. 371-401.

(13) Balls, M. et al. (1995), Practical aspects of the validation of toxicity test procedures. The report and

recommendations of ECVAM workshops, ATLA, Vol. 23, pp. 129-147.

TG 431 OECD/OCDE

14

© OECD, (2014)

(14) ICCVAM (Interagency Coordinating Committee on the Validation of Alternative Methods) (1997),

Validation and Regulatory Acceptance of Toxicological Test Methods, NIH Publication No. 97-

3981, National Institute of Environmental Health Sciences, Research Triangle Park, NC, USA.

http://iccvam.niehs.nih.gov/docs/guidelines/validate.pdf.

(15) ICCVAM (Interagency Coordinating Committee on the Validation of Alternative Methods) (2002),

ICCVAM evaluation of EpiDermTM

(EPI-200), EPISKINTM

(SM), and the Rat Skin Transcutaneous

Electrical Resistance (TER) assay: In Vitro test methods for assessing dermal corrosivity potential of

chemicals. NIH Publication No. 02-4502. National Toxicology Program Interagency Center for the

Evaluation of Alternative Toxicological Methods, National Institute of Environmental Health

Sciences, Research Triangle Park, NC, USA.

http://iccvam.niehs.nih.gov/methods/epiddocs/epis_brd.pdf

(16) EC-ECVAM (1998), Statement on the scientific validity of the EpiSkinTM

test (an in vitro test for

skin corrosivity), issued by the ECVAM Scientific Advisory Committee (ESAC10), 3 April 1998.

Available at: http://ecvam.jrc.ec.europa.eu.

(17) EC-ECVAM (2000), Statement on the application of the EpiDermTM

human skin model for skin

corrosivity testing, issued by the ECVAM Scientific Advisory Committee (ESAC14), 21 March

2000. Available at: http://ecvam.jrc.ec.europa.eu.

(18) Hoffmann, J. et al. (2005), Epidermal-skin-test 1000 (EST-1000)-A new reconstructed epidermis for

in vitro skin corrosivity testing, Toxicology in Vitro, Vol. 19/7, pp. 925-929.

(19) Kandárová, H. et al. (2006), Assessment of the human epidermis model SkinEthic RHE for in vitro

skin corrosion testing of chemicals according to new OECD TG 431, Toxicology in Vitro, Vol. 20/5,

pp. 547–559.

(20) Tornier, C., M. Roquet, A.B. Fraissinette (2010), Adaptation of the validated SkinEthicTM

Reconstructed Human Epidermis (RHE) skin corrosion test method to 0.5 cm2 tissue sample,

Toxicology in Vitro, Vol. 24/5, pp. 1379-1385.

(21) EC-ECVAM (2006), Statement on the application of the SkinEthicTM

human skin model for skin

corrosivity testing, issued by the ECVAM Scientific Advisory Committee (ESAC25), 17 November

2006. Available at: http://ecvam.jrc.ec.europa.eu.

(22) EC-ECVAM (2009), ESAC statement on the scientific validity of an in-vitro test method for skin

corrosivity testing: the EST-1000, issued by the ECVAM Scientific Advisory Committee (ESAC30),

12 June 2009. Available at: http://ecvam.jrc.ec.europa.eu.

(23) OECD (2013), “Summary Document on the Statistical Performance of Methods in OECD Test

Guideline 431 for Sub-categorisation”, OECD Environment, Health and Safety Publications (EHS),

Series on Testing and Assessment, No. 190, OECD Publications, Paris.

(24) Alépée, N., M.H. Grandidier, J. Cotovio (2014), Sub-categorisation of skin corrosive chemicals by

the EpiSkin™ reconstructed human epidermis skin corrosion test method according to UN GHS:

Revision of OECD Test Guideline 431, Toxicology in Vitro, Vol. 28/2, pp. 131-145.

OECD/OCDE TG 431

15

© OECD, (2014)

(25) Worth, A.P. et al. (1998), An Evaluation of the Proposed OECD Testing Strategy for Skin

Corrosion, ATLA, Vol. 26, pp. 709-720.

(26) Eskes, C. et al. (2012), Regulatory assessment of in vitro skin corrosion and irritation data within the

European framework: Workshop recommendations, Regulatory Toxicology and Pharmacology, Vol.

62/2, pp. 393-403.

(27) Mosmann, T. (1983), Rapid colorimetric assay for cellular growth and survival: application to

proliferation and cytotoxicity assays, Journal of Immunological Methods, Vol. 65/1, pp. 55-63.

(28) Tinois, E. et al. (1994), The Episkin model: Successful reconstruction of human epidermisin vitro, in

In vitro Skin Toxicology, Rougier, A., A.M. Goldberg, H.I. Maibach (eds.), Mary Ann Liebert, Inc.,

pp. 133-140.

(29) Cannon, C. L. et al. (1994), New epidermal model for dermal irritancy testing, Toxicology in Vitro,

Vol. 8/4, pp. 889 - 891.

(30) Ponec, M. et al. (2000), Lipid and ultrastructural characterization of reconstructed skin models,

International Journal of Pharmaceutics, Vol. 203/1-2, pp. 211 - 225.

(31) Tinois, E. et al. (1991), In vitro and post –transplantation differentiation of human keratinocytes

grown on the human type IV collagen film of a bilayered dermal substitute, Experimental Cell

Research, Vol. 193/2, pp. 310-319.

(32) Parenteau, N.L. et al. (1992), The organotypic culture of human skin keratinocytes and fibroblasts

to achieve form and function, Cytotechnology, Vol. 9, pp. 163-171.

(33) Wilkins, L.M. et al. (1994), Development of a bilayered living skin construct for clinical

applications, Biotechnology and Bioengineering, Vol. 43/8, pp. 47-756.

(34) EpiSkin™ SOP, INVITTOX Protocol No. 118 (December 2011), EpiSkin™ Skin Corrosivity Test.

Available at: http://ecvam.jrc.ec.europa.eu.

(35) EpiDerm™ SOP, Version MK-24-007-0024 (February 2012), Protocol for: In vitroEpiDerm™ skin

corrosion test (EPI-200-SCT), For use with MatTek Corporation's reconstructed human epidermal

model EpiDerm. Available at: http://ecvam.jrc.ec.europa.eu.

(36) SkinEthic™ RHE SOP, INVITTOX Protocol (January 2012), SkinEthicTM

Skin Corrosivity Test.

Available at: http://ecvam.jrc.ec.europa.eu.

(37) epiCS® SOP, Version 4.1 (January 2012), In Vitro Skin Corrosion: Human Skin Model Test

Epidermal Skin Test 1000 (epiCS® ) CellSystems. Available at: http://ecvam.jrc.ec.europa.eu.

TG 431 OECD/OCDE

16

© OECD, (2014)

ANNEX 1

PERFORMANCE STANDARDS FOR ASSESSMENT OF PROPOSED SIMILAR OR MODIFIED

IN VITRO RECONSTRUCTED HUMAN EPIDERMIS (RHE) TEST METHODS FOR SKIN

CORROSION2

INTRODUCTION

1. The purpose of Performance Standards (PS) is to provide the basis by which new or modified test

methods, both proprietary (i.e. copyrighted, trademarked, registered) and non-proprietary can demonstrate

to have sufficient reliability and relevance for specific testing purposes. The PS, based on valid and

accepted test methods, can be used to evaluate the reliability and relevance of other analogous test methods

(colloquially referred to as “me-too” test methods) that are based on similar scientific principles and

measure or predict the same biological or toxic effect (8). On the other hand, modified test methods, which

propose potential improvements to an approved test method, should be evaluated to determine the effect of

the proposed changes on the test method’s performance and the extent to which such changes affect the

information available for the other components of the validation process. Depending on the number and

nature of the proposed changes, the generated data and supporting documentation for those changes, they

should either be subjected to the same validation process as described for a new test method, or, if

appropriate, to a limited assessment of reliability and relevance using established PS (9).

2. Similar (me-too) or modified test methods proposed for use under this Test Guideline should be

evaluated to determine their reliability and relevance using Reference Substances (Table 1) representing

the full range of the TG 404 in vivo corrosivity scores, i.e., Corrosive (UN GHS Category 1A and Category

1B and 1C) and non-corrosive chemicals (1). The proposed similar or modified test methods should have

reliability and predictive capacity, which are comparable or better than those derived from the two VRM

[EpiSkinTM

(SM) and EpiDermTM

SCT (EPI-200)] and as described in paragraphs 6 to 10 of this Annex

(Tables 2 and 3) (11) (12) (24). The reliability of the new or modified test method, as well as its ability to

correctly identify non-corrosive and corrosive chemicals, and possibly also to discriminate UN GHS

Category 1A from Category 1B and 1C corrosive chemicals, should be determined prior to its use for

testing chemicals.

3. These PS are based on the US-ICCVAM PS (7) for evaluating the validity of new or modified

RhE test methods. The PS consists of (8): (i) essential test method components; (ii) recommended

Reference Substances, and; (iii) defined reliability and accuracy values that the proposed test method

should meet or exceed.

2Proposed new or modified test methods following the PS of this Test Guideline should be submitted to the OECD for

adoption and inclusion into the Test Guideline before being used for regulatory purposes.

OECD/OCDE TG 431

17

© OECD, (2014)

ESSENTIAL TEST METHOD COMPONENTS

4. These consist of essential structural, functional, and procedural elements of a validated test

method that should be included in the protocol of a proposed, mechanistically and functionally similar or

modified test method. These components include unique characteristics of the test method, critical

procedural details, and quality control measures. Adherence to essential test method components will help

to assure that a similar or modified proposed test method is based on the same concepts as the

corresponding VRMs (6). The essential test method components are described in detail in paragraphs 16 to

32 of the Test Guideline:

The general conditions (paragraph 16)

The functional conditions, which include:

- Viability (paragraph 17)

- Barrier function (paragraph 18)

- Morphology (paragraph 19)

- Reproducibility (paragraph 20)

- Quality control (paragraph 21)

The procedural conditions (paragraphs 22 to 32)

For specific parameters (e.g., for Tables 2, 3, 4, 5, and 6), adequate values should be provided for any new

similar or modified test method, these specific values may vary depending on the specific test method.

MINIMUM LIST OF REFERENCE SUBSTANCES

5. Reference Substances are used to determine if the reliability and relevance of a proposed similar

or modified test method, proven to be structurally and functionally sufficiently similar to the VRMs, or

representing a minor modification of one of the VRMs, are comparable or better than those of the VRMs

(11) (12) (24). The 30 recommended Reference Substances listed in Table 1 include substances

representing different chemical classes (i.e. chemical categories based on functional groups), and are

representative of the full range of TG 404 in vivo scores. The substances included in this list comprise 10

UN GHS Category 1A, 10 UN GHS Category 1B and 1C (the in vivo data do not permit distinction

between the two categories) and 10 non-corrosive substances. The substances listed in Table 1 are selected

from the substances used in the validation study of the VRMs, with regard to chemical functionality and

physical state (10) (11) (12) (24). These Reference Substances represent the minimum number of

chemicals that should be used to evaluate the reliability and relevance of a proposed similar or modified

test method able to discriminate between Category 1A, Category 1B and 1C and non-corrosive substances

and mixtures (1A vs. 1B and 1C vs. NC), in accordance with the UN GHS (1) . For similar or modified test

methods able to discriminate corrosive from non-corrosive substances and mixtures but not able to

sub-categorize corrosive chemicals (C vs. NC), only 20 of the 30 substances listed in Table 1 (the ones not

in italics) need to be evaluated (5 UN GHS Category 1A, 5 UN GHS Category 1B and 1C and 10 non-

corrosive substances). The use of these Reference Substances for the development/optimization of new

similar test methods should be avoided to the extent possible. In situations where a listed substance is

TG 431 OECD/OCDE

18

© OECD, (2014)

unavailable, other substances for which adequate in vivo reference data are available could be used,

primarily from the substances used in the validation study of the VRMs. If desired, additional substances

representing other chemical classes and for which adequate in vivo reference data are available may be

added to the minimum list of Reference Substances to further evaluate the accuracy of the proposed test

method.

OECD/OCDE TG 431

19

© OECD, (2014)

Table 1: Minimum list of Reference Substances for determination of Reliability and Predictive Capacity for

similar or modified in vitroRhE-based skin corrosion test methods. The 20 chemicals NOT in italics should be

tested with similar or modified test methods proposed to discriminate Corrosive from Non-Corrosive

chemicals (without sub-categorization). Additional reference substances should be tested with similar or

modified test methods proposed to identify Cat. 1A, a combination of Category 1B and 1C (referred to as

1B/1C below) and non-corrosive chemicals. These additional reference substances are indicated in italics.

Chemical1 CASR

N

Chemical

Class2

Physical

State

EpiSkinTM 4

EpiDermTM 4

SkinEthicTM

4

epiCS® 4

Non-corrosive chemicals based on in vivo results3

Phenethyl

bromide*

103-

63-9

Electrophile L (3) NC (3) NC (3) NC (2) NC

4-Amino-1,2,4-

triazole

584-

13-4

Organic base S (3) NC (3) NC (3) NC (2) NC

4-(methylthio)-

benzaldehyde*

3446-

89-7

Electrophile L (3) NC (3) NC (3) NC (2) NC

Lauric acid 143-

07-7

Organic acid S (3) NC

(3) NC (3) NC (2) NC

1,9-Decadiene 1647-

16-1

Neutral

organic

L (3) NC

(3) NC (3) NC (2) NC

2,4-

Dimethylaniline

95-68-

1

Organic base L (2) NC

(1) 1B/1C

(1) NC

(2) 1B/1C

(2) 1B/1C

(1) 1A

(1) NC

(1)

1B/1C

3,3-

Dithiopropionic

acid

1119-

62-6

Organic acid S (3) NC (3) NC (3) NC (2) NC

Methyl

palmitate

112-

39-0

Neutral

organic

S (3) NC

(3) NC (3) NC (2) NC

2-Hydroxyiso-

butyric acid

594-

61-6

Organic acid S (3) 1B/1C (3) 1B/1C (3) 1B/1C (2)

1B/1C

Sodium

undecylenate

(33%)

3398-

33-2

Soap /

Surfactant

L (3) 1B/1C (3) 1B/1C (3) 1B/1C (2)

1B/1C

UN GHS Cat. 1B-and-1C based on in vivo results3

Glyoxylic acid

monohydrate

563-

96-2

Organic acid S (3) 1B/1C (3) 1B/1C (3) 1B/1C (2)

1B/1C

Lactic acid 598-

82-3

Organic acid L (3) 1B/1C (3) 1B/1C (3) 1B/1C (2)

1B/1C

Sodium

bisulphate

monohydrate

10034-

88-5

Inorganic salt S (3) 1B/1C (3) 1B/1C (2) 1B/1C

(1) NC

(2)

1B/1C

Ethanolamine* 141-

43-5

Organic base Viscous (3) 1B/1C (3) 1B/1C (3) 1B/1C (2)

1B/1C

60/40

Octanoic/decan

oic acid

68937-

75-7

Organic acid L (3) 1B/1C (3) 1B/1C (3) 1B/1C (2)

1B/1C

Hydrochloric

acid (14.4%)

7647-

01-0

Inorganic acid L (3) 1B/1C (3) 1B/1C (3) 1B/1C (2)

1B/1C

Fluoroboric

acid

16872-

11-0

Inorganic acid L (3) 1A

(3) 1A (3) 1A (2) 1A

Propionic acid 79-09-

4

Organic acid L (3) 1A

(3) 1A (3) 1A (2) 1A

2-tert-

Butylphenol*

88-18-

6

Phenol L (3) 1B/1C (3) 1A (3) 1A (2) 1A

TG 431 OECD/OCDE

20

© OECD, (2014)

Cyclohexyl

amine*

108-

91-8

Organic base L (3) 1B/1C (3) 1A (3) 1A (2) 1A

UN GHS Cat. 1A based on in vivo results3

Acrylic acid 79-10-

7

Organic acid L (3) 1A

(3) 1A (3) 1A (2) 1A

Bromoacetic

acid

79-08-

3

Organic acid S (3) 1A

(3) 1A (3) 1A (2) 1A

Boron

trifluoride

dehydrate

13319-

75-0

Inorganic acid L (3) 1A (3) 1A (3) 1A (2) 1A

Phenol 108-

95-2

Phenol S (3) 1A

(3) 1A (3) 1A (2) 1A

Phosphorus

tribromide

7789-

60-8

Inorganic acid L (3) 1A (3) 1A (3) 1A (2) 1A

Silver nitrate 7761-

88-8

Inorganic salt S (1) 1A

(2) 1B/1C

(3) 1A (3) 1A (2) 1A

Formic acid 64-18-

6

Organic acid L (3) 1A

(3) 1A (3) 1A (2) 1A

Dichloroacetyl

chloride

79-36-

7

Electrophile L (3) 1A (3) 1A (3) 1A (2) 1A

Sulphuric acid

(98%)

7664-

93-9

Inorganic acid L (3) 1A (3) 1A (3) 1A

(2) 1A

N,N-Dimethyl

dipropylene

triamine*

10563-

29-8

Organic base L (3) 1B/1C (3) 1B/1C (3) 1B/1C (2)

1B/1C

Abbreviations: CASRN = Chemical Abstracts Service Registry Number; UN GHS = United Nations Globally

Harmonized System (1); NC = Not Corrosive 1These substances, sorted first by corrosives versus non-corrosives, then by corrosive sub-category, were selected

from the substances used in the ECVAM validation studies of EpiSkin™ and EpiDermTM

SCT (10) (11) (12) and

from post-validation studies based on data generated by EpiSkinTM

(24), EpiDermTM

, SkinEthicTM

and epiCS®

developers. Unless otherwise indicated, the substances were tested at the purity level obtained when purchased from a

commercial source (10) (12). The selection includes, to the extent possible, substances that: (i) are representative of

the range of corrosivity responses (e.g. non-corrosives; weak to strong corrosives) that the VRMs are capable of

measuring or predicting; (ii) are representative of the chemical classes used in the validation studies; (iii) reflect the

performance characteristics of the VRM; (iv) have chemical structures that are well-defined; (v) induce reproducible

results in the VRM; (vi) induce definitive results in the in vivo reference test method; (vii) are commercially

available; and (viii) are not associated with prohibitive disposal costs. Chemicals marked with an * are potential direct

MTT reducers. 2Chemical class assigned by Barratt et al. (1998) (10). 3The corresponding UN Packing groups are I, II and III, respectively, for the UN GHS 1A, 1B and 1C.

4The in vitro predictions reported in this table were obtained with the various test methods during post-validation

testing performed by the test method developers. These predictions were corrected for direct MTT reduction using killed

control tissues.

OECD/OCDE TG 431

21

© OECD, (2014)

DEFINED RELIABILITY AND ACCURACY VALUES

6. For purposes of establishing the reliability and relevance of proposed similar or modified RhE

test methods to be used by several independent laboratories, all 30 Reference Substances listed in Table 1

should be tested in at least three laboratories (24 Reference Substances for methods not able to sub-

categorize corrosive chemicals). It is however essential that all PS-based validation studies are

independently assessed by internationally recognized validation bodies, in agreement with international

guidelines (8). In each laboratory, all relevant Reference Substances should be tested for each exposure

time in three independent runs performed with different tissue batches and at sufficiently spaced time

points. Each run should consist of at least two concurrently tested tissue replicates per exposure time for

each test chemical, negative control, PC and adapted controls for direct MTT reduction and/or colour

interference.

7. The calculation of the reliability and predictive capacity of the proposed test method should be

done according to the rules described below to ensure that a predefined and consistent approach is used:

1. Within-laboratory reproducibility (WLR) should be calculated based on concordance of

classifications using only qualified tests obtained with Reference Substances for which at

least two qualified tests are available. In addition, it should be reported the number and

identity of the Reference Substances which per laboratory have none or only one qualified

test (omitted from WLR calculations), as well as how many and which Reference Substances

per laboratory have two or three qualified tests (used for WLR calculations).

2. For the calculation of between-laboratory reproducibility (BLR) the final classification for

each Reference Chemical in each participating laboratory should be obtained by using the

arithmetic mean value of viability over the different qualified tests performed. BLR should be

calculated based on concordance of classifications using only qualified tests from Reference

Substances for which at least one qualified test per laboratory is available. It should be

reported how many and which Reference Substances do not have at least one qualified test

per laboratory (omitted from BLR calculations), as well as how many and which Reference

Substances have 3, 4, 5, 6, 7, 8 or 9 qualified tests that can be used to calculate BLR (with at

least one qualified test per laboratory).

3. The calculation of predictive capacity (e.g. sensitivity, specificity and accuracy for C vs. NC)

should be done using all qualified tests obtained for each Reference Chemical in each

laboratory. The calculations should be based on the individual predictions of each qualified

test for each Reference Chemical in each laboratory and not on the arithmetic mean values of

viability over the different qualified tests performed.

In this context, a qualified test consists of a test that meets the criteria for an acceptable test, as defined in

the corresponding SOP, and is within a qualified run. Otherwise, the test is considered as non-qualified. A

qualified run consists of a run that meets the test acceptance criteria for the NC and PC, as defined in the

corresponding SOP. Otherwise, the run is considered as non-qualified.

Within-laboratory reproducibility

8. An assessment of within-laboratory reproducibility for similar or modified test method proposed

TG 431 OECD/OCDE

22

© OECD, (2014)

to discriminate corrosive from non-corrosive chemicals (but not to sub-categorize corrosive chemicals)

should show a concordance of classifications (corrosive or non-corrosive) obtained in different,

independent tests of the 24 relevant Reference Substances within one single laboratory equal or higher (≥)

than 90% (actual for EpiSkinTM

: 100%, 100% and 96% in each laboratory, respectively). An assessment of

within-laboratory reproducibility for similar or modified test method proposed to discriminate between

Category 1A, Category 1B and 1C and non-corrosive chemicals should show a concordance of

classifications (Category 1A, Category 1B and 1C or non-corrosive) obtained in different, independent

tests of the 30 Reference Substances within one single laboratory equal or higher (≥) than 80% (actual for

EpiSkinTM

: 96%, 96% and 88% in each laboratory, respectively).

Between-laboratory reproducibility

9. For similar or modified test methods proposed to discriminate corrosive from non-corrosive

chemicals (but not to sub-categorize corrosive chemicals), the concordance of classifications (corrosive or

non-corrosive) between a minimum of three laboratories, obtained for the 24 relevant Reference

Substances, should be equal or higher (≥) than 80% (actual for EpiSkinTM

: 88%). For similar or modified

test methods proposed to discriminate between Category 1A, Category 1B and 1C and non-corrosive

chemicals, the concordance of classifications (Category 1A, Category 1B and 1C or non-corrosive)

between a minimum of three laboratories, obtained for the 30 Reference Substances, should be equal or

higher (≥) than 70% (actual for EpiSkinTM

: 80%).

Predictive capacity

10. The predictive capacity of the proposed similar or modified test method should be comparable or

better to that of the VRMs. For similar or modified test method proposed to discriminate corrosive from

non-corrosive chemicals (C vs. NC) but not to sub-categorize corrosive chemicals, the sensitivity and

specificity obtained with the 20 relevant Reference Substances (Table 1) should be equal or higher (≥) than

95% and 70%, respectively, and the accuracy should be equal or higher (≥) than 82.5% (Table 2). For

similar or modified test method proposed to discriminate between Category 1A, Category 1B-and-1C and

non-corrosive chemicals (Cat. 1A vs. Cat. 1B-and-1C vs. NC), the minimum predictive capacity values

that should be obtained with the 30 Reference Substances (Table 1) are indicated in Table 3 for RhE test

methods similar to EpiSkin™ and for RhE test methods similar to EpiDerm™.

Table 2: Required sensitivity, specificity and accuracy for similar or modified RhE test methods to

be considered valid to discriminate corrosive from non-corrosive chemicals (C vs. NC) but not able

to sub-categorize corrosive chemicals. Values are based on the results of the two VRMs (EpiSkin™

and EpiDerm™) for the 20 Reference Substances not in italics from Table 1.

Sensitivity Specificity Accuracy

≥ 95%

(actual for EpiSkin™: 100%;

actual for EpiDerm™: 100%)

≥ 70%

(actual for EpiSkin™: 76.7%;

actual for EpiDerm™: 73.3%)

≥ 82.5%

(actual for EpiSkin™: 88.3%;

actual for EpiDerm™: 86.7%)

OECD/OCDE TG 431

23

© OECD, (2014)

Table 3: Required predictive capacity for similar or modified RhE test method to be considered

valid to discriminate between Cat. 1A, a combination of Category 1B and 1C (referred to as 1B-and-

1C below) and non-corrosive chemicals (Cat. 1A vs. Cat. 1B-and-1C vs. NC)*. Values are based on

the results of the two VRMs (EpiSkin™ and EpiDerm™) for the 30 Reference Substances (see table

1).

VRM EpiSkin™ EpiDerm™

Sensitivity

(C vs NC)

≥ 95%

(actual for EpiSkinTM

: 100.0%)

≥ 95%

(actual for EpiDermTM

: 100.0%)

Correctly classified 1A ≥ 80%

(actual for EpiSkinTM

: 83.3%)

≥ 90%

(actual for EpiDermTM

: 90.0%)

1A Underclassified 1B-

and-1C

≤ 20%

(actual for EpiSkinTM

: 16.7%,)

≤ 10%

(actual for EpiDermTM

: 10.0%,)

1A Underclassified NC 0%

(actual for EpiSkinTM

: 0.0%)

0%

(actual for EpiDermTM

: 0.0%)

Correctly classified 1B-

and-1C

≥ 80%

(actual for EpiSkinTM

: 80.0%)

≥ 55%

(actual for EpiDermTM

: 60.0%)

1B-and-1C

Overclassified 1A

≤ 20%

(actual for EpiSkinTM

: 20.0%)

≤ 45%

(actual for EpiDermTM

: 40.0%)

1B and 1C

Underclassified NC

≤ 5%

(actual for EpiSkinTM

: 0.0%)

≤ 5%

(actual for EpiDermTM

: 0.0%)

Specificity ≥ 70%

(actual for EpiSkinTM

: 76.7%)

≥ 70%

(actual for EpiDermTM

: 73.3%)

NC Overclassified 1A ≤ 5%

(actual for EpiSkinTM

: 0.0%)

≤ 5%

(actual for EpiDermTM

: 0.0%)

NC Overclassified 1B-

and-1C

≤ 30%

(actual for EpiSkinTM

: 23.3%)

≤ 30%

(actual for EpiDermTM

: 26.7%)

Accuracy

(C vs. NC)

≥ 87%

(actual for EpiSkinTM

: 92.2%)

≥ 87%

(actual for EpiDermTM

: 91.1%)

Accuracy

(1A vs. 1B-and-1C vs.

NC)

≥ 78%

(actual for EpiSkinTM

: 80.0%)

≥ 72%

(actual for EpiDermTM

: 74.4%)

* Depending on the results obtained with a similar or modified RhE test method for the 30 Reference Substances, it

may be considered similar to EpiSkin™ or similar to EpiDerm™ for the purpose of this Test Guideline. The

EpiSkinTM

and EpiDermTM

test methods are able to sub-categorize (i.e. 1A versus 1B-and-1C versus NC) but

differences are observed (SkinEthicTM

and epiCS®

are considered similar to EpiDerm™). For RhE test methods that

demonstrate similarity to EpiSkinTM

, results can be directly used based on the outcoming predictions. For RhE test

methods that demonstrate similarity to EpiDermTM

, chemicals that are classified as Category 1B-and-1C can be

considered as Category 1B-and-1C, whereas chemicals for which cell viability at 3 minutes is below 50% should be

considered as Category 1, since the Category 1A predictions of these three test methods contain a high rate of over-

predictions of chemicals of Categories 1B and 1C (see also paragraph 7 of the Test Guideline). The regulatory

framework in member countries will decide how this Test Guideline will be used, e.g. acknowledging the significant

probability of overclassification, a Category 1A classification may still be accepted or further testing may be

conducted to confirm the result.

Study Acceptance Criteria

11. It is possible that one or several tests pertaining to one or more test chemical and control

substance does/do not meet the test acceptance criteria or is/are not acceptable for other reasons (non-

TG 431 OECD/OCDE

24

© OECD, (2014)

qualified tests). To complement missing data, a maximum of two additional tests for each test chemical is

admissible per laboratory ("retesting"). More precisely, since in case of retesting also PC and NC have to

be concurrently tested, a maximum number of two additional runs may be conducted for each test chemical

in each laboratory. Importantly, each laboratory should not produce more than three qualified tests per test

chemical. Excess production of data and subsequent data selection are regarded as not appropriate.

12. It is conceivable that even after retesting, three qualified tests are not obtained for every

Reference Chemical in every participating laboratory, leading to an incomplete data matrix. In such cases

the following three criteria should all be met in order to consider the datasets acceptable for purposes of

PS-based validation studies:

1. All relevant Reference Substances (24 for Category 1 vs. Non Corrosive; 30 for Cat. 1A vs.

Cat. 1B and 1C vs. Non Corrosive) should have at least one complete test sequence in one

laboratory.

2. Each of the at least three participating laboratories should have a minimum of 85% complete

test sequences (for 24 Reference Substances: 3 incomplete test sequences are allowed per

laboratory; for 30 Reference Substances: 4 incomplete test sequences are allowed per

laboratory).

3. At least 90% of all test sequences from at least three laboratories need to be complete (for

24 Reference Substances tested in 3 laboratories: a total of 7 incomplete test sequences are

allowed; for 30 Reference Substances tested in 3 laboratories: a total of 9 incomplete test

sequences are allowed).

In this context, a test sequence consists of the total number of independent tests performed for a single

Reference Chemical in a single laboratory, including any re-testing (a total of 3 to 5 tests). A complete test

sequence consists of a test sequence containing three qualified tests. A test sequence containing less than

3 qualified tests is considered as incomplete.

OECD/OCDE TG 431

25

© OECD, (2014)

ANNEX 2

DEFINITIONS

Accuracy: The closeness of agreement between test method results and accepted reference values. It is a

measure of test method performance and one aspect of relevance. The term is often used interchangeably

with “concordance” to mean the proportion of correct outcomes of a test method (9).

Cell viability: Parameter measuring total activity of a cell population e.g. as ability of cellular

mitochondrial dehydrogenases to reduce the vital dye MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-

diphenyltetrazolium bromide, Thiazolyl blue), which depending on the endpoint measured and the test

design used, correlates with the total number and/or vitality of living cells.

Chemical: means a substance or a mixture

Complete test sequence: A test sequence containing three qualified tests. A test sequence containing less

than 3 qualified tests is considered as incomplete (see also definition of “test sequence” below).

Concordance: This is a measure of test method performance for test methods that give a categorical result,

and is one aspect of relevance. The term is sometimes used interchangeably with accuracy, and is defined

as the proportion of all chemicals tested that are correctly classified as positive or negative. Concordance is

highly dependent on the prevalence of positives in the types of test chemical being examined (9).

ET50: Can be estimated bydetermination of the exposure time required to reduce cell viability by 50%

upon application of the marker chemical at a specified, fixed concentration, see also IC50.

IC50: Can be estimated by determination of the concentration at which a marker chemical reduces the

viability of the tissues by 50% (IC50) after a fixed exposure time, see also ET50.

Infinite dose: Amount of test chemical applied to the epidermis exceeding the amount required to

completely and uniformly cover the epidermis surface.

Me-too test: A colloquial expression for a test method that is structurally and functionally similar to a

validated and accepted reference test method. Such a test method would be a candidate for catch-up

validation. Interchangeably used with similar test method (9).

Mixture: means a mixture or solution composed of two or more substances in which they do not react.

MTT: 3-[4,5-dimethylthiazol-2-yl]-2, 5-diphenyltetrazolium bromide)

NC: Non corrosive

TG 431 OECD/OCDE

26

© OECD, (2014)

OD: Optical Density

PC: Positive Control

Performance standards (PS): Standards, based on a validated test method, that provide a basis for

evaluating the comparability of a proposed test method that is mechanistically and functionally similar.

Included are; (i) essential test method components; (ii) a minimum list of Reference Substances selected

from among the chemicals used to demonstrate the acceptable performance of the validated test method;

and (iii) the similar levels of reliability and accuracy, based on what was obtained for the validated test

method, that the proposed test method should demonstrate when evaluated using the minimum list of

Reference Substances (9).

Qualified run: A run that meets the test acceptance criteria for the NC and PC, as defined in the

corresponding SOP. Otherwise, the run is considered as non-qualified.

Qualified test: A test that meets the criteria for an acceptable test, as defined in the corresponding SOP,

and is within a qualified run. Otherwise, the test is considered as non-qualified.

Reference Substances: Chemicals selected for use in the validation process, for which responses in the in

vitro or in vivo reference test system or the species of interest are already known. These chemicals should

be representative of the classes of chemicals for which the test method is expected to be used, and should

represent the full range of responses that may be expected from the chemicals for which it may be used,

from strong, to weak, to negative. Different sets of Reference Substances may be required for the different

stages of the validation process, and for different test methods and test uses (9).

Relevance: Description of relationship of the test method to the effect of interest and whether it is

meaningful and useful for a particular purpose. It is the extent to which the test method correctly measures

or predicts the biological effect of interest. Relevance incorporates consideration of the accuracy

(concordance) of a test method (9).

Reliability: Measures of the extent that a test method can be performed reproducibly within and between

laboratories over time, when performed using the same protocol. It is assessed by calculating intra- and

inter-laboratory reproducibility (9).

Run: A run consists of one or more test chemicals tested concurrently with a negative control and with a

PC.

Sensitivity: The proportion of all positive/active chemicals that are correctly classified by the test method.

It is a measure of accuracy for a test method that produces categorical results, and is an important

consideration in assessing the relevance of a test method (9).

Skin corrosion in vivo: The production of irreversible damage of the skin; namely, visible necrosis

through the epidermis and into the dermis, following the application of a test chemical for up to four hours.

Corrosive reactions are typified by ulcers, bleeding, bloody scabs, and, by the end of observation at

OECD/OCDE TG 431

27

© OECD, (2014)

14 days, by discoloration due to blanching of the skin, complete areas of alopecia, and scars.

Histopathology should be considered to evaluate questionable lesions.

Specificity: The proportion of all negative/inactive chemicals that are correctly classified by the test

method. It is a measure of accuracy for a test method that produces categorical results and is an important

consideration in assessing the relevance of a test method (9).

Substance: means chemical elements and their compounds in the natural state or obtained by any

production process, including any additive necessary to preserve the stability of the product and any

impurities deriving from the process used, but excluding any solvent which may be separated without

affecting the stability of the substance or changing its composition.

Test: A single test substance concurrently tested in a minimum of two tissue replicates as defined in the

corresponding SOP.

Test sequence: The total number of independent tests performed for a single test substance in a single

laboratory, including any re-testing. A test sequence may include both qualified and non-qualified tests.

Test chemical: means what is being tested

Tiered testing strategy: Testing which uses test methods in a sequential manner; the test methods selected

in each succeeding level are determined by the results in the previous level of testing (9).

UN GHS (United Nations Globally Harmonized System of Classification and Labelling of

Chemicals): A system proposing the classification of chemicals (substances and mixtures) according to

standardized types and levels of physical, health and environmental hazards, and addressing corresponding

communication elements, such as pictograms, signal words, hazard statements, precautionary statements

and safety data sheets, so that to convey information on their adverse effects with a view to protect people

(including employers, workers, transporters, consumers and emergency responders) and the environment

(1).

TG 431 OECD/OCDE

© OECD, (2014)

28

ANNEX 3

MAIN TEST METHOD COMPONENTS OF THE RhE TEST METHODS VALIDATED FOR SKIN CORROSION TESTING

Test Method

Components

EpiSkinTM

EpiDermTM

SCT SkinEthicTM

RHE epiCS®

Model surface 0.38 cm2 0.63 cm

2 0.5 cm

2 0.6 cm

2

Number of

tissue replicates At least 2 per exposure time 2-3 per exposure time At least 2 per exposure time

At least 2 per exposure time

Treatment

doses and

application

Liquids and viscous: 50 µL ± 3 µL

(131.6 µL/cm2)

Solids: 20 2 mg (52.6 mg/cm2) + 100

µL ± 5µL NaCl solution (9 g/L)

Waxy/sticky: 50 2 mg (131.6

mg/cm2) with a nylon mesh

Liquids: 50 µL (79.4 µL/cm2) with or

without a nylon mesh

Pre-test compatibility of test chemical

with nylon mesh

Semisolids: 50 µL (79.4 µL/cm2)

Solids: 25 µL H2O (or more if

necessary) + 25 mg (39.7 mg/cm2)

Waxes: flat “disc like” piece of ca. 8

mm diameter placed atop the tissue

wetted with 15 µL H2O.

Liquids and viscous: 40 µL ± 3µl (80

µL/cm2) using nylon mesh

Pre-test compatibility of test chemical

with nylon mesh

Solids: 20 µL ± 2µl H2O + 20 3 mg

(40 mg/cm2)

Waxy/sticky: 20 3 mg (40 mg/cm2)

using nylon mesh

Liquids: 50 µL (83.3 µL/cm2) using

nylon mesh

Pre-test compatibility of test chemical

with nylon mesh

Semisolids: 50 µL (83.3 µL/cm2)

Solids: 25 mg (41.7 mg/cm2) + 25 µL

H2O (or more if necessary)

Waxy: flat “cookie like” piece of ca. 8

mm diameter placed atop the tissue

wetted with 15 µL H2O

Pre-check for

direct MTT

reduction

50 µL (liquid) or 20 mg (solid)

+ 2 mL MTT

0.3 mg/mL solution for 180 5 min

at 37oC, 5% CO2, 95% RH

if solution turns blue/purple, water-

killed adapted controls should be

performed

50 µL (liquid) or 25 mg (solid)

+ 1 mL MTT

1 mg/mL solution for 60 min

at 37oC, 5% CO2, 95% RH

if solution turns blue/purple, freeze-

killed adapted controls should be

performed

40 µL (liquid) or 20 mg (solid)

+ 1 mL MTT

1 mg/mL solution for 180± 15 min at

37°C, 5% CO2, 95% RH

if solution turns blue/purple,

freeze-killed adapted controls should

be performed

50 µL (liquid) or 25 mg (solid)

+ 1 mL MTT

1 mg/mL solution for 60 min

at 37oC, 5% CO2, 95% RH

if solution turns blue/purple,

freeze-killed adapted controls should

be performed

OECD/OCDE TG 431

© OECD, (2014)

29

Test Method

Components

EpiSkinTM

EpiDermTM

SCT SkinEthicTM

RHE epiCS®

Pre-check

for colour

interference

10 µL (liquid) or 10 mg (solid) + 90 µL

H2O mixed for 15 min at RT

if solution becomes coloured, living

adapted controls should be

performed

50 µL (liquid) or 25 mg (solid) + 300

µL H2O

for 60 min at 37oC, 5% CO2, 95% RH

if solution becomes coloured, living

adapted controls should be

performed

40 µL (liquid) or 20mg (solid) + 300

µL H2O mixed for 60

min at RT

if test chemical is coloured, living

adapted controls should

be performed

50 µL (liquid) or 25 mg (solid) + 300

µL H2O

for 60 min at 37oC, 5% CO2, 95% RH

if solution becomes coloured,

living adapted controls

should be performed

Exposure

time and

temperature

3 min, 60 min ( 5 min)

and 240 min ( 10 min)

In ventilated cabinet

Room Temperature (RT, 18-28oC)

3 min at RT, and 60 min

at 37oC, 5% CO2, 95% RH

3 min at RT, and 60 min

at 37oC, 5% CO2, 95% RH

3 min at RT, and 60 min

at 37oC, 5% CO2, 95% RH

Rinsing 25 mL 1x PBS (2 mL/throwing) 20 times with a constant soft stream

of 1x PBS

20 times with a constant soft stream

of 1x PBS

20 times with a constant soft stream

of 1x PBS

Negative

control

50 µL NaCl solution (9 g/L)

Tested with every exposure time

50 µL H2O

Tested with every exposure time

40 µL H2O

Tested with every exposure time

50 µL H2O

Tested with every exposure time

Positive control 50 µL Glacial acetic acid

Tested only for 4 hours

50 µL 8N KOH

Tested with every exposure time

40 µL 8N KOH

Tested only for 1 hour

50 µL 8N KOH

Tested with every exposure time

MTT solution 2 mL 0.3 mg/mL 300 µL 1 mg/mL 300 µL 1 mg/mL 300 µL 1 mg/mL

TG 431 OECD/OCDE

© OECD, (2014)

30

Test Method

Components

EpiSkinTM

EpiDermTM

SCT SkinEthicTM

RHE epiCS®

MTT

incubation

time and

temperature

180 min ( 15 min) at 37oC, 5% CO2,

95% RH 180 min at 37

oC, 5% CO2, 95% RH

180 min (± 15 min) at 37oC, 5% CO2,

95% RH 180 min at 37

oC, 5% CO2, 95% RH

Extraction

solvent

500 µL acidified isopropanol

(0.04 N HCl in isopropanol)

(isolated tissue fully immersed)

2 mL isopropanol

(extraction from top and bottom of

insert)

1.5 mL isopropanol

(extraction from top and bottom of

insert)

2 mL isopropanol

(extraction from top and bottom of

insert)

Extraction time

and

temperature

Overnight at RT, protected from light

Overnight without shaking at RT or for

120 min with shaking (~120 rpm) at

RT

Overnight without shaking at RT or

for 120 min with shaking (~120

rpm) at RT

Overnight without shaking at RT or

for 120 min with shaking (~120

rpm) at RT

OD reading 570 nm (545 - 595 nm)

without reference filter

570 nm (or 540 nm)

without reference filter

570 nm (540 - 600 nm)

without reference filter

540 - 570 nm

without reference filter

Tissue Quality

Control

18 hours treatment with SDS

1.0 mg/mL ≤ IC50 ≤ 3.0 mg/mL

Treatment with 1% Triton X-100

4.08 hours ≤ ET50 ≤ 8.7 hours

Treatment with 1% Triton X-100

4.0 hours ≤ ET50 ≤ 10.0 hours

Treatment with 1% Triton X-100

2.0 hours ≤ ET50 ≤ 7.0 hours

OECD/OCDE TG 431

© OECD, (2014)

31

Test Method

Components

EpiSkinTM

EpiDermTM

SCT SkinEthicTM

RHE epiCS®

Acceptability

Criteria

1. Mean OD of the tissue replicates

treated with the negative control

(NaCl) should be ≥ 0.6 and ≤ 1.5 for

every exposure time

2. Mean viability of the tissue

replicates exposed for 4 hours with

the positive control (glacial acetic

acid), expressed as % of the negative

control, should be ≤ 20%

3. In the range 20-100% viability and

for ODs≥ 0.3, difference of viability

between the two tissue replicates

should not exceed 30%.

1. Mean OD of the tissue replicates

treated with the negative control

(H2O) should be ≥ 0.8 and ≤ 2.8 for

every exposure time

2. Mean viability of the tissue

replicates exposed for 1 hour with the

positive control (8N KOH), expressed

as % of the negative control, should

be < 15%

3. In the range 20 - 100% viability, the

Coefficient of Variation (CV)

between tissue replicates should be

30%

1. Mean OD of the tissue replicates

treated with the negative control

(H2O) should be ≥ 0.8 and ≤ 3.0 for

every exposure time

2. Mean viability of the tissue

replicates exposed for 1 hour (and 4

hours, if applicable) with the

positive control (8N KOH),

expressed as % of the negative

control, should be 15%

3. In the range 20-100% viability, and

for ODs ≥ 0.3, difference of viability

between the two tissue replicates

should not exceed 30%

1. Mean OD of the tissue replicates

treated with the negative control

(H2O) should be ≥ 0.8 and ≤ 2.8 for

every exposure time

2. Mean viability of the tissue

replicates exposed for 1 hour with

the positive control (8N KOH),

expressed as % of the negative

control, should be 20%

3. In the range 20-100% viability, and

for ODs ≥ 0.3, difference of viability

between the two tissue replicates

should not exceed 30%

TG 431 OECD/OCDE

© OECD, (2014)

32

ANNEX 4

PERFORMANCE OF TEST METHODS FOR SUB-CATEGORISATION

This annex provides a table where performances of the four test methods were calculated based on a

set of 80 chemicals tested by the four test developers. Calculations were performed by the OECD

Secretariat, reviewed and agreed by an expert subgroup (23).

EpiSkinTM

, EpiDermTM

, SkinEthicTM

and epiCS® test methods are able to sub-categorize (i.e. 1A

versus 1B-and-1C versus NC) but differences are observed between EpiSkinTM

and the three other test

methods, EpiDerm™, SkinEthic™ and epiCS®, for sub-categorization. Results from EpiSkin

TM can be

directly used based on the outcoming results, whereas results from EpiDermTM

, SkinEthicTM

and epiCS®,

should take into account high over-classification rates from those three test methods for 1B-and-1C

category (see table 4.0 in Annex 4). Therefore, for EpiDermTM

, SkinEthicTM

and epiCS®, chemicals that are

classified as 1B-and-1C can be considered as 1B-and-1C, and chemicals for which cell viability at 3

minutes is below 50% should be considered as 1, that is to say that either under the prediction principle

they could be claimed as 1A or they should undergo further testing to be possibly confirmed as 1B-and-1C.

The regulatory framework in member countries will decide how this Test Guideline will be used.

OECD/OCDE TG 431

© OECD, (2014)

33

Table 4.0: Performances, Overclassification rates, Underclassification rates, and Accuracy

(Predictive capacity) of the four test methods based on a set of 80 chemicals all tested over 2 or 3

runs in each test method.

NC=non corrosive.

STATISTICS ON ENTIRE SET OF CHEMICALS

(n= 80 chemicals tested over 2 or 3 runs, i.e. 159* or 240 classifications) *one chemical was tested once because of no availability

EpiSkinTM

EpiDermTM

SkinEthicTM

epiCS®

Overclassifications:

Cat. 1BC chemicals that are overclassified 1A

21.50% 41.94% 46.24% 45.90%

Cat. NC chemicals that are overclassified 1B/1C

20.72% 23.42% 24.32% 28.38%

Cat. NC chemicals that are overclassified 1A

0.00% 2.70% 2.70% 0.00%

Cat. NC chemicals that are overclassified Corr.

20.72% 26.13% 27.03% 28.38%

Global overclassification rate (all categories)

17.92% 28.33% 30.42% 30.82%

Underclassifications:

Cat. 1A Underclassified 1B/1C

16.67% 8.33% 13.89% 8.33%

Cat. 1A Underclassified NC

0.00% 0.00% 0.00% 0.00%

Cat. 1BC Underclassified NC

2.15% 0.00% 7.53% 6.56%

Global Underclassification rate (all categories)

3.33% 2.47% 5.00% 3.77%

Correct Classifications:

1A Correctly classified

83.33% 91.67% 86.11% 91.67%

1B/1C Correctly classified

76.34% 58.06% 46.24% 47.54%

NC Correctly classified

79.28% 73.87% 72.97% 71.62%

Accuracy (Predictive capacity)

78.75%

70.42%

64.58% 65.41%